christinminor459/OnionClaw — reverse-engineered prompt
Reverse engineered prompt
Build me a Python tool called OnionClaw that gives an AI agent safe, practical access to the Tor network for legitimate OSINT and threat intelligence work. I want it to work two ways, as a simple standalone terminal app and as a drop in OpenClaw skill with almost no setup beyond an env file.
It should let me verify Tor is running, rotate to a new Tor identity, check a bunch of dark web search engines, search across them at once with deduplicated results, fetch both onion and normal pages through Tor, and run AI analysis on scraped content in a few different investigation modes. I also want one end to end pipeline command that can take a query, refine it, search, filter, scrape, and summarize the findings.
Please keep it zero fuss, readable, and usable by someone doing security research. Assume Tor is running locally and make the AI analysis optional if no key is set. Add basic docs and examples, and look up current docs online if you need to.
Want more depth? Deep Reverse